J. Org. Chem., Vol. 42, No. 11,1977 2023
Notes Second-order rate constants were determined under pseudoM, excess of first-order conditions, Brz concentration = 5 X unsaturated substrate varied from 160to 20 (for phenylpropiolic acid due to low solubility in water only tenfold excess of substrate was used). The second-orderrate constant of trans-cinnamicacid in water was determined at comparable concentrations of both substrates, both being about 5 X lo-* m and containing 1%of CH30H to ensure solubility. For runs carried out on a Cary 16 spectrophotometer,a 10-cm cell was used to obtain the absorption change of ca. 0.2 absorbance unit. Acknowledgment. Continued financial support by the National Research Council of Canada is gratefully acknowledged. We also thank Professor G. Modena, Istituto di Chimica Organica, Universita di Padova, Padova, Italy, for permission to use his data. Registry No.-Styrene,
A
B
C
D
E
Figure 1. IH NMR spectrum (Varian T-60) of the CH2 protons in 1-methyl-3-phospholene1-oxide at various dilutions; A, neat; B, 0.2 parts; C, 0.3 parts; D, 0.5 parts; E, 1 part of CDC13 (v/v).
100-42-5; phenylacetylene, 536-74-3;
trans-cinnamic acid, 140-10-3; sodium cinnamate, 18509-03-0;
phenylpropiolic acid, 637-44-5;sodium phenylpropiolate, 7063-23-2; p-nitrocinnamic acid, 619-89-6;p-nitrophenylpropiolic acid, 221624-2. References a n d Notes (1) R. E. Buckles, J. M. Bader, and R . J. Thurmaier, J. Org. Chem., 27, 4523 (1962). (2) R . C. Fahey and H. J. Schneider, J. Am. Chem. SOC.,80, 4429 (1968). (3) G. Heublein, J. Prakt. Chem., 31, 84 (1966). (4) K . Yates, G. ti. Schmid. T. W. Regulski, D. G. Garratt, H. W. Leung, and R . McDonald, J. Am. Chem. SOC.,95, 160 (1973). (5) J. R. Ruasse and J. E. Dubois, J. Org. Chem., 87, 1977 (1975). (6) J. E. Dubois and X. 0.Huynh, Bull. SOC. Chirn. Fr., 1468 (1968). (7) J. E. Dubois and H. Herzog, Bull. SOC.Chim. Fr., 57 (1968). (8) K. B. Harbison, Ph.D. Thesis, Massachusetts Institute of Technology, 1966. (9) K. Yates, R . S.McDonald, and S.A. Shapiro, J. Org. Chem., 38, 2460 (1973). (10) J. A . Pincock and K. Yates. Can. J. Chem., 48, 3332 (1970).
Unusual Shielding Effects in the Proton Nuclear Magnetic Resonance Spectrum of 1-Methyl-3-phospholene 1-Oxide
100 Hz
Figure 2. Experimental (bottom) and computer-simulated (top) 270-MHz lH NMR spectra of the CHz protons in 1-methyl-3-phospholene 1-oxide;A, neat (at 50 "C); B, 50%solution in CDC13 at room temperature; C, 9%solution in CDC13 at room temperature.
Kurt Moedritzer*and Pierre A. Berger Corporate Research Department, Monsanto Company, St. Louis, Missouri 63166 Received December 7,1976 In the course of a study of the chemical and physical properties of phospholene d e r i ~ a t i v e s l we - ~ were puzzled by the l H NMR spectrum of one of the compounds of this class, 1methyl-3-phospholene 1-oxide3 (I). In agreement with a
spectrum of the phospholene derived from I, obtained by reduction of the tertiary phosphine oxide I to the corresponding tertiary phosphine, did show nonequivalent protons5 HA and HB and thus suggests an AA'BB'X pattern (X = 31P),no explanation was advanced for the inconsistency in the l H NMR spectrum of I. Nonequivalence of the methylene protons was also observed for the sulfide1 derived from I and for the 1-chloro- and 1hydroxy-3-phospholene l - ~ u l f i d e sIn . ~ view of these observations and of the known rigidity of the stereochemistry around the phosphorus atom, the lH NMR pattern obtained for I was difficult to rationalize. We, therefore, undertook a more detailed study of concentration and temperature effects on the 'H NMR spectra of the methylene protons of I. Results a n d Discussion
I previous report5 we found that a solution of I in CDCls showed a simple 60-MHz 'H NMR pattern consisting of three doublets (due to coupling with 31P),for CH3 at d 1.60 ( 2 J ~ c=p 13 Hz), CH2 at d 2.43 ( V ~ c p = 11 Hz), and CH at d 5.87 ppm ( 3 J ~ c c=p 28 Hz), suggesting a high degree of symmetry for the molecule. From such spectra it was concluded earlies that protons HAand HB in I fail to show the nonequivalence expected of protons in cis.and in trans position to the P-0 bond, relative to the plane of the ring. Although the l H NMR
Proton spectra of I recorded a t room temperature and a t 60 MHz in the neat state and a t various degrees of dilution in CDC13 are shown in Figure 1. Surprisingly, with increasing dilution, the initially complex methylene proton spectrum simplifies to the doublet reported earlier.5 This effect is also shown upon dilution with benzene as solvent. Two possible explanations suggest themselves for this observation: (a) a dynamical effect, which renders the A and B protons equivalent within the NMR time scale, or (b) an accidental simplification of an AA'BB'X spectrum. In order to shed more light on this problem proton spectra were obtained a t 270 MHz as shown in Figure 2 for the following concen-
2024 J.Org. Chem., ?lol. 42, No. 11,1977 Table I. NMR Parametersa (in Hertz at 270 MHz) of the CH2 Protons in I-Methyl-3-phospholene I-Oxide Neat bA
I2
- h3 J
I
~
~
2
J
~
~
2
J
~
~
-53.2 17.4 f 14.32 =7.95
9% solution (CDCl3)
+35.56 17.64 f16.04? ~9.18~
a The olefinic protons He give rise to a clean doublet due to coupling with 31P and show no evidence of coupling with the methylene protons HI\ and HB. Absolute shifts: HA,2.59; HB, 2.47 ppm. HAin I is cis to P-0 in accordance with J and b assignments made for 1,2,ti-trimethyl-3-phospholene1-oxide.7
Notes parameters were not obtained because of the insufficiency of actual experimental values provided by the spectrum in Figure 2. These observations show clearly that the methylene protons A and B of I are nonequivalent, basically giving rise to an AA’BB’ spectrum, which degenerates, under proper conditions of the environment (dilution, temperature), into apparent ABX or even A& spectra. The temperature dependence of a neat sample a t 60 MHz is shown in Figure 3. With increasing temperature the spectrum becomes simpler and more symmetrical, in a way similar to the effect observed upon dilution a t 60 MHz. The underlying reason for this is, as in the dilution study, not a dynamical but a shielding effect. We believe that the reason for these observations, based solely on the present NMR study, is a matter of speculation. It is, however, plausible to suggest that various degrees of molecular association or short-range molecular ordering are the basis of these interesting shielding effects. Acknowledgment. We wish to thank Professor W. H. Urry, University of Chicago, for the 270-MHz spectra and stimulating discussions, and R. E. Miller for experimental work. Registry No.-I, 930-38-1. References a n d Notes (1) (2) (3) (4) (5) (6) (7)
K. Moedritzer, Synth. React. Inorg. Met.-Org. Chem., 4, 119 (1974). K. Moedritzer, Synth. React. Inorg. Met-Org. Cbem., 5, 45 (1975). K. Moedritzer, Synth. React. Inorg. Met.-Org. Chem., 5, 299 (1975). K. Moedritzer, 2.Naturforsch. B., 31, 709 (1976). L. D. Ouin, J. P, Gratz, and T. P. Earket, J. Org. Chem., 33, 1034 (1968). D. Gagnaire, J. E. Robert, and J. Verrier, Chem. Commun., 819 (1967). A. Bond, M. Green, and S. C. Pearson, J. Chem. Soc. B, 929 (1968).
A Novel Synthesis of Trifluoromethylthioacetic Acid R. M. DeMarinis* and W. M. Bryan Research & Development Division, S m i t h Kline &French Laboratories, Philadelphia, Pennsylvania 19101 Received December 7.1976
Figure 3. lH NMR spectrum (Varian A-50/60) of the CH2 protons in a neat sample of 1-methyl-3-phospholene1-oxideat various tem-
peratures. trations: neat sample I, !SO and 9% solutions of I in CDCls. The spectrum of the neat sample had to be recorded at 50 OC because of the difficulty of preventing the sample from crystallizing a t room temperature. The others were recorded a t room temperature. In view of the temperature effect discussed below, the spectrum of the neat sample is, therefore, not directly comparable with the corresponding 60-MHz spectrum. The spectra of the neat compound and of the 9% solution, however, are satisfactorily interpreted in terms of the AB part of ABX spectra,6 with the parameters summarized in Table I. Computer-simulated spectra of I for the two cases of Table I are also shown in Figure 2. While the coupling constants are only slightly affected (to within 12%) by dilution, the chemical shifts reflect an inversion in the relative shielding experienced by A and B. The spectrum of the 50% solution in Figure 2 cannot be understood as an AB part of an ABX pattern; however, computer simulations revealed that it could qualitatively be explained as the AA’BB’ part of an AA’BB’X pattern. Exact
The synthesis of trifluoromethylthioacetic acid,1!2an important intermediate in the preparation of the semisynthetic cephalosporin antibiotic cefazaflur (SKF E19962),~is made difficult by the limited methods available for the elaboration of the trifluoromethylthio moiety) We wish to report a facile route to this compound from ethyl mercaptoacetate which provides a potentially useful method for the conversion of thiols to trifluoromethyl sulfides. Chlorination of a methyl group adjacent to sulfur followed by halogen exchange with an inorganic fluoride a t elevated temperature is one of the most widely use4 methods for the synthesis of a trifluoromethylthio group.5 Thus, our initial approach to the synthesis of trifluoromethylthioacetic acid was to prepare the corresponding trichloride and convert it by existing methods to the trifluoride. However, in methylthioacetic acid the methylene adjacent to the sulfur is sufficiently activated by the carboxyl group to preclude selective chlorination of the methyl. As an alternate approach we tried fluorination of the readily available disulfide 1, followed by sulfur extrusion to give an ester of trifluoromethylthioacetic acid. Although the fluoride exchange could not be carried out by the usual method, Le., antimony trifluoride-antimony pentachloride, we discovered that it occurred rapidly under mild conditions using phase transfer catalysis. Treatment of a hexane solution of 1, pre-
